steam to enter with the primary air, the clinkers are kept softer, are not so injurious to the furnace walls, and a valuable heating gas is produced in the generator thereby. The steam under the grate bars should be regulated at about 30 pounds of water per 100 pounds of coke consumed. The furnace door should never be opened during clinkering. To raise the heat, inspect the air and gas nostril holes, and if in order prick the fuel on the grate bars to allow the free passage of primary air. Open the dampers slightly and if, after one or two charges, there should be no material change, the primary, secondary and main dampers should be opened % of an inch at a time and the results carefully noted. When the temperature of the bench is too low a considerable amount of liquid tar will be found in the mouthpiece. To lower the heat, slightly close the dampers, or in connection with this the primary air slides, and regulate the secondary slides to give the proper combustion. The heat in the recuperators should not be more than a dull red below the secondary air inlets, as this will probably mean too little secondary air. The draught (a) under the grate bars, (b) above the coke, (c) at the entrance of gases into the recuperator, and (d) at the outlet of the recuperator, should be measured and noted on paper once in every watch. The CO canals, combustion chamber, and entrance to the chimneys are to be inspected several times on each watch. The CO canals should never be white hot, and only short blue flames should enter the chimneys. If the blue flames are long, secondary air is wanting. If no blue flames are noticeable, more or less secondary air is in excess. If bright yellow flames are present, it is the sign of a leaky retort. All these observations must be made when not clinkering. Test the benches daily for leaks with a lighted torch. All pressure gauges are to be kept in first-class condition. A vacuum of 6-10 is sufficient at the outlet of the last waste gas flue, and is obtained with from 400° to 500° F. A supply of sand and a shovel should be kept on the stack for use in case of accident. Fresh mud should also be kept on hand in the retort house. Forty-two inches of coke in the generator is sufficient, although 50 inches is better. The air regulators and dampers, which should not be constantly altered, are to be placed so that the interior of the bench seems slightly veiled in fog. Flames in the bench show too little air, and a perfectly clear interior too much air. Carbonic acid gas determinations should be made once a week, and are most easily obtained with a special apparatus designed. for this purpose by Mr. William von Oechelhaeuser, of Dessau, Germany. A sketch of this is shown in the following diagram, the operation being as follows: The vessel B is charged with a solution consisting of 1⁄2 ounce or 16 gr. of caustic potash or caustic soda dissolved in enough distilled water to fill it to the middle of C. The gas is then forced or drawn through the upper bulb A from c to a until a good sample is secured. Both these openings are then closed, the whole apparatus inverted a few times, placed on a table, and the siphon gauge attached, from which the percentage of carbon dioxide can be directly read. If the temperature before and after the absorption differs, 1-3 per cent. must be added or subtracted for every degree of increase or decrease of temperature respectively. The carbonic acid gas in the generator gases should not exceed 3 per cent.; in the chimney it should reach a maximum of 19 per cent. (at least 17 per cent.), which will show that there is no uncombined or superfluous oxygen in the waste gases. If less than 19 per cent. carbonic acid is found, two causes are possible; there is either unburnt CO in the waste gases, which will only heat the chimney, or a surplus of oxygen (air) is in combination. Both these defects can not occur at once, and to determine which is present a test for CO must be made. If CO is not present then oxygen is in excess in as many per cent. as the CO was less than 19 per cent. The excess equals five times the excess of oxygen. Shutting Down.-After the charges, which should be left in the retorts, are burnt, close the stack valve, seal up the primary and secondary air slides and other openings with clay, and when the bench is cool and dark open the ascension pipe caps fully. Dip pipes that have a cap in common with the ascension pipes should be closed with wooden stoppers to prevent dust from falling into the hydraulic main. Draw the charges and loose retort carbon, clean the furnace, oil the iron work where necessary, pull out the stoppers, and lay the retort lids on loosely. It will be seen from the foregoing, that the value of a recuperative furnace depends on: I. Practical mechanical construction. 2. Even and equal distribution of heat. 3. Amount of fuel necessary for firing, per 100 pounds of coal carbonized. 4. Small loss of draught from a point under the grate bars to the outlet of recuperator. 5. Minimum amount of CO2 in generator gases. 6. Maximum amount of CO: in chimney gases. DISCUSSION. The President called upon Mr. Frederic Egner to open the dis cussion. MR. F. EGNER-Mr. Steinwedell is to be complimented on his choice of a subject. He has not followed in the wake of the recently somewhat popular fad of writing for a technical Association a paper on the commercial aspect of the gas business, and making an easy "try" for cheap applause by an account of bizarre methods to increase gas sales. He has gone to the beginning of the matter the retort house-wherein, as one of the fathers of gas lighting, Samuel Clegg, said, "The money is made." But while Mr. Steinwedell has treated his subject in a reasonably thorough and instructive manner, we may be excused if we ask for general information, a question or two; and thus perhaps bring out a few additional facts suggested by his admirable paper. It would be interesting, no doubt, if Mr. Steinwedell would explain why, if the recuperative furnaces save 50 per cent. in fuel, 40 per cent. in wear and tear of the retorts, 30 per cent. in floor space per 1,000 cubic feet of gas made, and 25 per cent. in labor, they would not do for any gas works which may be classed as such, regardless of cost of fuel or additional outlay in cost of construction. It seems to me that Mr. Steinwedell has been quite conservative in his statements as to the savings which can be affected by employing the recuperative furnace. In some respects his figures are much too low. For instance, the actual physical labor saved the fireman if not the company, in all cases, is fully twice 25 per cent. And if we save 40 per cent. in wear and tear of the bench by using that style of a furnace, a low cost of fuel should not deter anyone from having a recuperative bench, not to mention the other advantages very properly enumerated by Mr. Steinwedell. Perhaps when writing his paper he had in mind conditions such as we find sometimes, where, from the nature of the ground, we can not go down to construct a pit, or where the cost of going up, may be raising the whole retort house in order that a stage floor could be introduced, would be prohibitive. While Mr. Steinwedell has given the exact relative conditions which govern the saving to be expected by the employment of recuperative furnaces, either full or half depth, he has not stated, probably supposing it to be generally understood, that management, intelligent or otherwise, will affect the results obtained from either, so that sometimes we find that one man gets better results from a half-depth furnace than another does from a well designed, well constructed full-depth furnace, although Mr. Steinwedell clearly shows that the full-depth furnace ought to give the best results. Mr. Steinwedell has clearly, and correctly, though concisely, enunciated the principles which govern the proper management of the recuperative furnace, and which if closely followed must result in successful operations. He, therefore, deserves the thanks of all those who have not yet had any experience with that kind of a furnace, and also the same from some who have, and who may have found snags in the way of smooth sailing with them. Mr. Steinwedell says that "The steam under the grate bars should be regulated at about 30 pounds of water to the 100 pounds of coke consumed." And it would be interesting to know how he advises to accomplish that. Does he favor a reliance altogether upon the water evaporated under the grate bars in the ash pan of the furnace, or would he favor admitting steam obtained from other sources by means of a convenient pipe suitably arranged? I think I would favor the latter course as the more reliable. Mr. Steinwedell's direction to frequently inspect the various parts of the bench during each watch, and in other parts daily, is a very good one to follow, if you can get men to do it; and his remarks upon the best depth of fuel to be maintained in a recuperative furnace are worthy of special notice. Many times have I seen furnaces at one end of which the depth of fuel would be 60 to 72 inches, while at the opposite end of the same furnace the depth would not be 30 inches at any time, and nothing at all sometimes, with results in accordance, viz.: Irregular heats in the bench, cracked and sagging retorts, stopped ascension pipes, large coke consumption, heavy, hard clinker in the furnace; overworked (through their own laziness and ignorance) stokers, and irregular yield, both as to quality and quantity of gas produced. This single feature of proper depth of fuel in the furnace which Mr. Steinwedell mentions is alone one, if acted upon, which makes this paper especially worthy to have been brought before us. It is this even depth of fuel, more easily kept in some constructions than in others, which has been instrumental in recommending the so-called "Munich furnace" to many of the successful and able gas engineers in the East, though why it should be called the "Munich furnace" and not just as well the "St. Louis furnace" it will not be so easy to explain, for we used that kind of a furnace at the old Laclede gas works some years before the same became |